Acta Pædiatrica ISSN 0803–5253

REGULAR ARTICLE

Prevalence of gestational exposure to cannabis in a Mediterranean city bymeconium analysisJaime Lozano1,2, Oscar Garcıa-Algar1,2, Emilia Marchei3, Oriol Vall1,2, Toni Monleon4, Rita Di Giovannandrea3, Simona Pichini (simona.pichini@iss.it)3

1.Unitat de Recerca Infancia i Entorn (URIE), Paediatric Service, Hospital del Mar, Barcelona, Spain2.Departament de Pediatria, Ginecologia i Obstetricia, i Medicina Preventiva, Universitat Autonoma, Barcelona, Spain3.Department of Therapeutic Research and Medicines Evaluation, Istituto Superiore di Sanita, Rome Italy4.Departament of Statistics, Universitat de Barcelona, Spain

KeywordsCannabis use, Pregnancy, Meconium analysis

CorrespondenceSimona Pichini, Department of TherapeuticResearch and Medicines Evaluation, IstitutoSuperiore di Sanita, V.le Regina Elena 299,00161, Rome, Italy.Tel: + 39-06-49903682 |Fax: + 39-06-49902016 |Email: simona.pichini@iss.it

Received12 May 2007; revised 12 July 2007;accepted 24 August 2007.

DOI:10.1111/j.1651-2227.2007.00535.x

AbstractAim: To investigate the prevalence of in utero exposure to cannabis in a cohort of 974 mother-infants

dyads from Barcelona (Spain).

Methods: A validated gas chromatographic-mass spectrometric assay was used for meconium

analysis.

Results: Prenatal exposure to cannabis was detected in 5.3% newborns; however, only 1.7% of the

participating mothers disclosed gestational drug use. Ethnicity was not associated with cannabis use,

nor were the maternal features or newborns somatometry. A significantly higher percentage of

cannabis using mothers had a managerial professional job versus nonusers.

Conclusions: Meconium analysis is especially important to identify neonates who have been exposed to cannabis

in utero, which appear normal at birth and therefore may not be recognized. Timely detection of these newborns

at risk provides the basis for appropriate treatment and adequate medical and social follow-up.

INTRODUCTIONThere is a large body of evidence that meconium analysiscan be used to assess the prevalence of in utero exposureto drugs of abuse during the second and third trimestersof gestation (1). Large-scale epidemiological studies basedon meconium analysis in newborns have been conductedin the United States (1,2). Only two studies on the preva-lence of illicit drugs use during pregnancy using meconiumtoxicology have been performed in Europe (3,4). One studyin the city of Glasgow (UK) analyzed 400 meconium sam-ples of newborns from a population with a high percent-age of social deprivation and drug misuse. Cannabinoidswere detected in 13.25% samples, cocaine in 2.75% and am-phetamine in 1.75% (3). The other study, called ‘MeconiumProject’ included a cohort of 1151 women from an urbanarea of Barcelona (Spain) with a high percentage (more than40%) of immigrants. Prenatal exposure to opiates, cocaineand combined drugs exposure in newborns was 8.7, 4.4 and2.2% respectively, with only one case of ecstasy consumption(4).

Using the same cohort, we investigated the prevalence ofcannabis consumption during gestation and tried to assessthe characteristics of the drug using pregnant mothers andtheir newborns using self-reported questionnaire as well asmeconium analysis of �9-tetrahydrocannabinol (THC) andits principal metabolites as drug biomarkers in this biologicalmatrix.

This study was supported by Dipartimento Nazionale per le politicheantidroga della Presidenza del Consiglio dei Ministri, Roma (Italy).

METHODSThe population of mother-infants dyads was recruited as pre-viously described (4). The study was approved by the Institu-tional Ethical Committee (CEIC-IMAS) in accordance withthe Declaration of Helsinki. At the time of delivery, pregnantwomen signed an informed consent, underwent a completeclinical examination and a structured interview. All new-borns were clinically examined at birth and somatometrydata as well as clinical signs were recorded.

Meconium (the total amount contained in the diaper) wascollected at 24 h after delivery, and immediately stored in dif-ferent aliquots at −20◦C until analysis. Of the 1151 samplesinitially obtained from the ‘Meconium Project’ cohort andpreviously analyzed for opiates, cocaine and amphetamines(4), aliquots to perform analysis of THC and its metabo-lites were available in 974 (84.6%). The 974 mother-infantsdyads did not show any statistical difference (p > 0.2) insocio-demographic characteristics and newborn somatome-try with respect to the 177 excluded dyads (data not shown).A validated gas chromatographic-mass spectrometric assaypreviously described (5) was used to quantify THC, 11-hydroxy-�9-tetrahydrocannabinol (THC-OH) and 11-nor-�9-tetrahydrocannabinol-9-carboxylic acid (THC-COOH).Meconium samples were analyzed within 1 year; there wasno degradation of the analytes under investigation duringthis time period (5).

Dyads with meconium negative for drugs of abuse werecompared with dyads with meconium positive for cannabis,opiates and cocaine. Statistical analysis for the continuousvariables was made with the Student’s t-test for two groupsand the analysis of variance (ANOVA) for three or moregroups. The Fisher’s exact test was used for the comparison

1734 C©2007 The Author(s)/Journal Compilation C©2007 Foundation Acta Pædiatrica/Acta Pædiatrica 2007 96, pp. 1734–1737

Lozano et al. Gestational exposure to canabis and meconium analysis

Table 1 Presence of cannabis biomarkers in the 52 positive samples

Each specific biomarker in all the samplesTHC n (%) 5 (9.6%)THC-OH n (%) 36 (69.2%)

THC-COOH n (%) 42 (80.8%)

Each combination in all the samplesTHC n (%) 3 (5.8%)THC-OH n (%) 7 (13.5%)THC-COOH n (%) 12 (23.1%)THC, THC-OH n (%) 0 (0%)THC, THC-COOH n (%) 1 (1.9%)THC-OH, THC-COOH n (%) 28 (53.8%)THC, THC-OH, THC-COOH n (%) 1 (1.9%)

of categorical data. For the analysis of individual percent-ages of exposed infants versus unexposed, the binomial testwas used. Multifactorial ANOVA was performed to assessthe relationship between in utero exposure to cannabis andbirth weight, gestational age and tobacco exposure. Statisti-cal significance was set at p < 0.05. The SPSS (version 12.0,SPSS, Inc., Chicago, USA) statistical package was used forthe analysis of data.

RESULTSOf the 974 samples, 52 (5.3%) were positive for cannabiswith 8 of them suggesting polydrug exposure since 2 werealso positives for cocaine, 5 for opiates and 1 for both co-caine and opiates in addition to cannabis.

Cannabis consumption at any time of pregnancy wasself-reported by 17 mothers (1.7% of total cohort), 11 ofwhich were confirmed by meconium analysis. The remaining6 mothers declared consumption only in the first trimesterof gestation and accordingly the meconiums were negativefor cannabis.

As shown in Table 1, THC-COOH alone was thebiomarker most frequently found in positive samples(80.8%). None of the combination of different biomarkersshowed higher percentage of positivity.

The paternal and maternal socio-economic and demo-graphic characteristics of the 876 dyads negative to any il-licit drug and the 52 dyads positive for cannabis are shownin Table 2. The remaining 46 dyads, whose meconium waspositive to other illicit drugs but cannabis, were not includedin the analysis.

A significantly higher percentage of cannabis using moth-ers had a managerial professional job compared to nonus-ing mothers, who in turn held unskilled jobs by a highpercentage.

Conversely, although there were a great number of immi-grants in the study cohort (more than 40%), ethnicity wasnot associated with drug use, nor were the maternal fea-tures or the newborns somatometry, even after adjusting forneonate or maternal characteristics.

DISCUSSIONMeconium analysis revealed significant levels of cannabisuse during pregnancy which was under-reported by the

mothers, as it was already shown in case of opiates and co-caine in our study cohort (4).

As has been described with other illicit drugs in thesame cohort (4), the objectively measured 5.3% preva-lence of cannabis use by pregnant women can be consid-ered extremely high if compared with self-reported decla-ration. Furthermore, interestingly, the prevalence disclosedby meconium analysis was very similar to the 4.7% youngSpanish nonpregnant women (15–34 years of age) admittingto weekly cannabis consumption in the last National Surveyon Drug Abuse (6). With respect to these data, our estima-tion of cannabis consumption during pregnancy might evenbe conservative. Indeed, since meconium formation startsbetween the 12th and 16th week of gestation, its analysisdoes not detect drug use during the first trimester of preg-nancy.

Similar to the gestational use of opiates and cocaine (4),it could be argued that the prevalence of cannabis use isdue to mainly, lower socio-economic status. Nonetheless,a significantly higher percentage of cannabis users had amanagerial professional job with respect to nonusers, whichwere statistically more represented in the lowest profes-sional category. This was not the case of 8.7% opiatesor the 4.4% cocaine consumption in the same study co-hort, which was not associated with maternal profession(4).

Even in the presence of high gestational consumption ofcocaine and opiates, polydrug use in cannabis using moth-ers was objectively assessed by meconium analysis only in15% samples. Furthermore, in contrast to opiates and co-caine use during pregnancy (4), gestational cannabis wasnot associated with a higher percentage of active tobaccosmoking.

Newborn somatometry was not affected by maternal con-sumption of cannabis, even after adjusting for tobacco smok-ing, neonate sex, maternal age and gestation weeks. This isin agreement with previous reports in large-scale, prospec-tive drug screening of newborns by meconium analysis (1),and with the fact that poorer somatometry in newborns pre-natally exposed to drugs of abuse is mainly explained by thehighest maternal smoking associated with drug consumptionduring pregnancy.

Our findings suggest that within this study cohort, the5.3% gestational cannabis use occurred mainly in a socialenvironment different from that of the 7.9% opiates and co-caine using pregnant women. From this observation, we canconclude that meconium analysis highlighted an overall hid-den consumption of drugs of abuse in more than 10% of ourstudy population.

Maternal screening for illicit drugs would be advisableto disclose drug use during pregnancy, but blood or urineanalysis can only detect for acute use and must be repeatedvery often. In spite of the fact that some authors cautionedagainst the use of meconium analysis of illicit drugs to esti-mate the degree of exposure during pregnancy (1), we shouldrecognize the importance of this particular diagnostic toolfor purposes of public policy. Although we agree that theeffort and expense involved in collecting and conducting

C©2007 The Author(s)/Journal Compilation C©2007 Foundation Acta Pædiatrica/Acta Pædiatrica 2007 96, pp. 1734–1737 1735

Gestational exposure to canabis and meconium analysis Lozano et al.

Table 2 Characteristics of the mother-infant dyads according to meconium analysis

Meconium negative for all drugs Meconium positive for cannabis onlyn = 876 n = 52 (5.3%)

Mother’s nationality (%)Spanish 50.0 47.7 (p = 0.91)European (other than Spanish) 6.4 6.8 (p = 0.74)American 22.8 27.3 (p = 0.39)African 10.8 6.8 (p = 0.60)Asian 10.0 11.4 (p = 0.68)

Father’s nationality (%)Spanish 49.5 41.5 (p = 0.32)European (other than Spanish) 6.2 2.4 (p = 0.49)American 22.9 24.4 (p = 0.83)African 10.6 17.1 (p = 0.28)Asian 10.8 14.6 (p = 0.50)Mother’s Job (Yes/No) (%)Yes 47.7 51.4 (p = 0.67)

Mother’s profession (%)Managerial professional [I] 5.5 22.2 (p = 0.03)∗

Skilled (nonmanual) [II] 13.1 0.0 (p = 0.13 Fisher1)Skilled (manual) [III] 22.1 16.7 (p = 0.76)Partly skilled [IV] 39.3 50.0 (p = 0.39)Unskilled [V] 20.0 11.1 (p = 0.00)∗∗∗

Father’s Job (Yes/No) (%)Yes 91.8 93.1 (p = 1.00)

Father’s profession (%)Managerial professional [I] 7.0 11.1 (p = 0.43)Skilled (nonmanual) [II] 4.3 7.5 (p = 0.35)Skilled (manual) [III] 12.0 14.8 (p = 0.75)Partly skilled [IV] 55.0 55.5 (p = 0.95)Unskilled [V] 21.7 11.1 (p = 0.31)

Maternal age (years) mean (SD) 29.3 (6.2) 28.9 (6.4) (p = 0.66)

Previous pregnancies (%)0 55.2 51.0 (p = 0.56)1 29.1 28.6 (p = 0.94)>2 15.7 20.4 (p = 0.42)

Previous premature infants (%)0 95.3 95.9 (p = 1.00)1 4.5 2.0 (p = 0.71)>2 0.2 2.0 (p = 0.16)

Previous abortions (%)0 64.9 59.2 (p = 0.42)1 26.0 24.5 (p = 0.81)>2 9.1 16.3 (p = 0.18)

Smoking during pregnancy (Yes/No) (%)Yes 20.5 26.7 (p = 0.33)Number of daily cigarettes mean (SD) 11.1 (8.6) 11.8 (9.9) (p = 0.64)

Newborn somatometryNewborn weight, mean (SD) 3260.6 (507.4) 3307.2 (531.2) p = 0.54Newborn height, mean (SD) 49.2 (2.3) 49.1 (2.0) p = 0.73Newborn cranial perimeter, mean (SD) 34.2 (1.5) 34.4 (1.5) p = 0.35

Sex (%)Male 50.4 56.0 (p = 0.42)

Premature (%)<37 weeks 7.6 0 (p = 0.04∗)≥37 weeks 92.4 100 (p = 0.04∗)

Somatometry adjusted for cigarette smoking, sex, maternal age and gestation weeksNewborn weight, mean (CI 95%) 3212.4 (3174.1–3250.8) 3281.7 (3108.9–3454.4)Newborn height, mean (CI 95%) 48.8 (48.6–49.0) 48.9 (48.3–49.6)Newborn cranial perimeter, mean (CI 95%) 34.1 (33.9–34.2) 34.3 (33.8–34.8)

1(p = 0.000 applying Binomial test normal approximation. However, normal approximation may be inaccurate for small samples).

1736 C©2007 The Author(s)/Journal Compilation C©2007 Foundation Acta Pædiatrica/Acta Pædiatrica 2007 96, pp. 1734–1737

Lozano et al. Gestational exposure to canabis and meconium analysis

meconium analysis is high, investigations in our study co-hort demonstrate that the methodology could be and shouldbe simplified. Indeed, as already shown for opiates and co-caine users, where measurement of only 6-monoacetyl mor-phine and cocaine in meconium accounted for almost 90%of the positive samples (4), 80% of the gestational cannabisuse could be recognized by measuring a single biomarker,THC-COOH.

Furthermore, if prenatal exposure to opiates and co-caine can also be recognized by other diagnostic tools(neonatal abstinence syndromes, poor newborn somatom-etry, less reluctance by addicted mothers to admit druguse during pregnancy) meconium analysis can be espe-cially important to identify pregnant cannabis consumersand neonates who have been exposed to cannabis inutero, which appear normal at birth and may not beidentified.

CONCLUSIONSConsidering the consequences of prenatal exposure tocannabis on neurobehavioral development (7), improved de-tection of newborns at risk provides the basis for appro-priate treatment and adequate medical and social follow-up. The identification of socio-economic and demographicalcharacteristics of the women at risk of cannabis consump-tion during pregnancy can help in addressing and imple-menting health policy interventions towards this group ofpeople.

ACKNOWLEDGEMENTSThe authors thank Marta Pulido, M.D., for editing themanuscript and editorial assistance.

References

1. Ostrea EM Jr, Brady M, Gause S, Raymundo AL, Stevens M.Drug screening of newborns by meconium analysis: alarge-scale, prospective, epidemiologic study. Pediatrics 1992;89: 107–13.

2. Lester BM, ElSohly M, Wright LL, Smeriglio VL, Verter J,Bauer CR, et al. The maternal lifestyle study: drug use bymeconium toxicology and maternal self-report. Pediatrics 2001;107: 309–17.

3. Williamson S, Jackson L, Skeoch C, Azzim G, Anderson R.Determination of the prevalence of drug misuse by meconiumanalysis. Arch Dis Child Fetal Neonatal Ed 2006; 91: F291–2.

4. Pichini S, Puig C, Zuccaro P, Marchei E, Pellegrini M, Murillo J,et al. Assessment of exposure to opiates and cocaine duringpregnancy in a Mediterranean city: preliminary results of the“Meconium Project”. Forensic Sci Int 2005; 153: 59–65.

5. Marchei E, Pellegrini M, Pacifici R, Palmi I, Lozano J,Garcia-Algar O, et al. Quantification of Delta9-tetrahydrocannabinol and its major metabolites in meconiumby gas chromatographic-mass spectrometric assay: assayvalidation and preliminary results of the “Meconium Project”.Ther Drug Monit 2006; 28: 700–6.

6. http://www.pnsd.msc.es/Categoria2/publica/pdf/oed-2004.pdf(March 2007).

7. Fried PA, Smith AM. A literature review of the consequences ofprenatal marihuana exposure. An emerging theme of adeficiency in aspects of executive function. Neurotoxicol Teratol2001; 23: 1–11.

C©2007 The Author(s)/Journal Compilation C©2007 Foundation Acta Pædiatrica/Acta Pædiatrica 2007 96, pp. 1734–1737 1737